Browse Prior Art Database

Method for Easy Statistical View of Storage Disk Technology in Storage Servers and Virtulized Storage Farms Disclosure Number: IPCOM000188346D
Original Publication Date: 2009-Oct-01
Included in the Prior Art Database: 2009-Oct-01
Document File: 3 page(s) / 102K

Publishing Venue



Disclosed is a solution which helps storage administrators with highly consumable and easy statistical view of storage disk technology in storage servers and virtualized storage farms. The solution covers its applicability to server based storage as well as virtualized clustered storage accessed via clustered filesystems like IBM General Parallel File Systems .

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 51% of the total text.

Page 1 of 3

Method for Easy Statistical View of Storage Disk Technology in Storage Servers and Virtulized Storage Farms

With the advent of Flash (SSD) and its advantages over (Hard Disk Drive) HDD, hybrid storage have started becoming common. Also in future with the advent of Phase Change memory and Racetrack technology based disk, the hybrid nature of storages will not only persist but grow. Each of these disk vary in its characteristics, features, costs etc depending upon its underlying technology and smart administrators

judiciously make use of these disks. For example, flash based disks come with faster

seek time but are costlier compared to HDD - so its availability in storage box may be limited and the administrator will judiciously use the flash disk by keeping only the most frequently used data.

Problem Part 1:

An enterprise like IBM* has many test labs. Each lab has hundreds of servers stacked up in number of racks to provide massive computing power similar to data centers. Storage is essential element of any server. As various storage technologies are being introduced, the server machines will soon be having multiple storage technology type disks attached to them. Servers with hybrid storage will start becoming common. So the problem will arise if the administrator wants to keep track of storage utilization of each storage technology on every server machines. This process worsens if there are hundreds of machines. Optionally the administrator may use the storage manager software available. This method will involve client/server software installation on every server and still the administrator will need to log in onto the central administration node. Additionally if the network fails, this method is of no use. Moreover the solution is targeting consumability where the administration at a glace can get a snapshot of the available storage per disk's technology per server. So if the administration comes to know that the FLASH based disk is getting full, while there is space for Hard Disk Drive (HDD) and Phase-change he can plan to procurer SSD in his limited budget.

Problem Part 2

In scale out storages, the storages are managed via clustered filesystems like IBM General Parallel File Systems** (GPFS***) which span across different storage nodes. So in hybrid storages, the filesystem will cluster across HDD drives , Flash Drives, Phase-Change Drives or even Racetrack drives. The current problem is Clustered Parallel Filesystems like GPFS sees these various storage technologies simply as an I/O device. The I/O device may be HDD, SDD or Phase Track. But for GPFS it is nothing but a NSD (Network Shared Disk). Consider a typical storage requirement of an enterprise. They have GPFS setup with hundreds of HDDs, a few SSDs and a couple of Phase Track devices. These devices will simply get listed by a different NSD name under...